HW4

library(plotly)
library(dplyr)
library('data.table')
library("stringr") 

d <- read.csv("Properties_philly_Kraggle_v2.csv")
d <- d[ d$Sale.Date != ('') , ] 
d <- d %>% mutate(Sale.Price.bid.price = trimws(Sale.Price.bid.price, whitespace = "[$]*\\s*")) %>% 
  filter(nzchar(Sale.Price.bid.price))
d[c("Price1","Price2")] <- str_split_fixed(d$Sale.Price.bid.price, ',', 2)
d$Price <- paste(d$Price1, d$Price2, sep="")
d$Price <-as.numeric(gsub(",", ".", d$Price))

fig <- d %>%
  plot_ly(
    x = ~PropType,
    y = ~Price,
    name = 'Actual Price',
    side = 'positive',
    type = 'violin',
    box = list(
      visible = T
    ),
    meanline = list(
      visible = T
    ),color = I("#B7D6C6")
  ) 


fig <- fig %>%
  add_trace(
    x = ~PropType,
    y = ~Opening.Bid,
    name = 'Opening Bid',
    side = 'negative',
    box = list(
      visible = T
    ),
    meanline = list(
      visible = T
    ),color = I("#CF909B")) 

fig %>% layout(title = 'Comparison Opening Bid to Actual Price')